In South America, Facebook and Google Drive 70% of Traffic

So far in 2016, in North America, real-time entertainment apps drive data demand on fixed networks, largely Netflix traffic, which represented about 35 percent of all bits transferred over North American fixed networks, according to Sandvine.

Amazon Video drove about 4.3 percent of peak downstream traffic.

On North American mobile networks, music services as a whole are increasing in traffic share, yet no single service is among the top 10 applications.

During peak period, real-time entertainment traffic continues to be by far the most dominant traffic category on mobile networks, accounting for almost 40 percent of the downstream bytes on the network.

In Latin American mobile networks, Facebook and Google drive 70 percent of total traffic in the region.

10% of Wearable Owners Have Stopped Using Them

You never can be too sure what the “next big thing” will be, in the connected devices business.

Tablets and smartwatches might be among the categories that have failed to achieve that status. People use them, to be sure. But neither device yet has yet to prove it is transformative in the same way that PCs earlier, and smartphones recently, have been.

Some 10 percent of wearable owners have stopped using their wearable, according to an Ericsson ConsumerLab poll of wearable device owners from Brazil, China, South Korea, the United Kingdom and United States.

Ericsson says about 33 percent of those who abandoned use of wearables did so within the first few weeks of ownership.

Limited functionality was the primary problem for 21 percent of respondents who stopped using their wearable device. Some 14 percent of of those who had stopped using their wearable device said it was because the devices were not a standalone product, and required use of a smartphone.

source: Ericsson

IT Shifting from "Cloud First" to "Cloud Only," Gartner Says

By 2019, more than 30 percent of the 100 largest vendors' new software investments will have shifted from cloud-first to cloud-only, Gartner analysts now predict.

"More leading-edge IT capabilities will be available only in the cloud, forcing reluctant organizations closer to cloud adoption,” said Yefim V. Natis, Gartner VP.

By 2020, more compute power will have been sold by IaaS (infrastructure as a service) and PaaS (platform as a service) cloud providers than sold and deployed into enterprise data centers.

The Infrastructure as a Service (IaaS) market has been growing more than 40 percent in revenue per year since 2011, and it is projected to continue to grow more than 25 percent per year through 2019.

By 2019, the majority of virtual machines (VMs) will be delivered by IaaS providers. By 2020, the revenue for compute IaaS andPlatform as a Service (PaaS) will exceed $55 billion — and likely pass the revenue for servers.

“Unless very small, most enterprises will continue to have an on-premises (or hosted) data center capability,” said Thomas Bittman, Gartner VP.

In Every Ecosystem, One Segment's Cost is Another Segment's Revenue

Ecosystems always are contentious, since one segment’s revenue is another segment’s cost.

We saw an obvious example of that when Verizon’s fixed network workers went on strike for higher wages and benefits.

We see the conflict often when various parties argue about rights to use spectrum; whether spectrum can be shared; whether spectrum should be available license exempt, or not; whether zero rating should be allowed.

We also will see that principle in action as the Indian government auctions a prodigious amount of mobile spectrum--as much as 2300 MHz-- later in 2016. Consider that the whole Indian mobile industry presently uses between 200 MHz and 300 MHz of spectrum. So the upcoming auction represents an order of magnitude (10 times) increase in spectrum.  

To put that into perspective, the potential spectrum rights could cost more than double the industry's gross revenue and more than 20 times the annual free cash flow of the entire mobile industry.

The expected spectrum payments also represent a sum four times higher than the last auction.

Shockingly, projected spectrum payments in this one auction could represent a sum as high as twice as much spending as in all prior mobile spectrum auctions put together.

So there you have a clear example of ecosystem tension because one segment’s revenue is another segment’s cost. The Indian government thinks it could raise a sum representing as much as 25 percent of the government’s total annual revenues.  

Some analysts would note that India already has some of the highest costs in the world, where it comes to spectrum rights.

In the end, all costs, everywhere in the full ecosystem, are paid by other parts of the ecosystem, with all ultimate costs borne by end users and consumers, or parties subsidizing use of ecosystem products (advertisers, for example).

Uber Requires Smartphones; Airbnb Arguably Does Not

It is probably not too early to speculate that smartphones now have been around long enough that people see new ways to leverage the technology. More accurately, people are coming to understand how apps, running on smartphones that people always have with them, are creating new possibilities.

Without smartphones, it is unlikely that Uber and other ridesharing services would be feasible. Airbnb likely would be conceivable, even without smartphones. But that obviously raises a question: in what other spheres of economic activity might smartphone-based apps be used to bring latent resources to market?

Let us be clear. People using smartphones to comparison shop, while they are out and about, and then ordering online, already is changing the face of retailing. But Uber is more profound than that.

What Uber did was revolutionize our thinking about the real-time use of assets that mostly lie fallow, even more than Airbnb did. Airbnb allows people to bring latent lodging assets to market, though not necessarily as a smartphone-essential function.

Uber really does require the ubiquity of smartphones to really demonstrate value. So the only question is what other areas of life have latent resources not brought to market because of friction, including the lack of a marketplace to monetize those assets.

Uber could not function without smartphones. Airbnb arguably can do so. So which potential spheres of economic life are smartphone-dependent, where it comes to commercializing latent assets that are underused?

"Impossible to Predict" How Consumers Will React to a Truly New Product, AWS Head Says

"What we've learned over time at Amazon, is that very often at the beginning of something really different and really new, it's impossible to predict how customers are going to react to the offering," says Andy Jassy, Amazon Web Services CEO.

That's why Amazon, as a company, doesn't focus on the first wave of feedback to any product,

Few firms have enough financial depth, or patience, to do so. But that approach resembles what Apple used to do, under Steve Jobs. Famously, Jobs believed consumers could not really judge whether they would like a new product they never had experienced, so market research was useless.

Access Providers Literally Cannot Own and Control IoT, So Partner

If current forecasts are correct, it literally will be impossible for access providers to dominate or control the broad Internet of Things business.

That might suggest it makes more sense to invest in stakes throughout the rest of the ecosystem--partnering, in other words--rather than staking everything on owning and controlling sizable portions of the value chain.

The Internet of Things, in all its forms--has a total potential economic impact of $3.9 trillion to $11.1 trillion a year by 2025, according to an analysis by McKinsey Global Institute. The biggest segments might well be industrial automation, smart cities and health apps.

At the top end, that level of value—including the consumer surplus—would be equivalent to about 11 percent of the world economy.

With the caveat that today’s access providers likely always will earn a majority of their revenue providing access services of various types (mobile, fixed, Internet access, Wi-Fi hotspot access,

Ethernet and other high-bandwidth connections), the key to future revenue is likely going to come from providing managed apps and services, as always has been the case.

That statement should be unremarkable, except for the fundamental change in business model since the advent of the Internet and IP communications.

In the past, one might have argued that most revenue came from selling apps (voice and messaging), and relatively less from “data access” connections (dumb pipe T-1, DS3 and so forth).

Even data access services such as frame relay, MPLS and ATM essentially were managed services, or “apps.”

These days, as apps are structurally separated from access, legacy communications providers are earning less revenue from fully-owned apps and lots more from data access (dumb pipe).

So the issue is how the revenue model might evolve over the next decade or two. Broadly speaking, it is clear that dumb pipe access and apps now are separate parts of the revenue stream. So the big question is how access providers can create viable and useful roles for themselves in the app part of the business.

Entertainment video is an early example of diversification. Other telco forays into new revenue sources have not been nearly that successful. Oddly enough, it is cable TV operators who have developed a model that should work for telcos, if telcos can create both the right internal mindset and consistently show a willingness to “make money without owning everything.”

The emphasis there is “making money,” not “owning everything.” All too often, new service initiatives seem to fail because the approach is too heavy handed and too controlling.

There is another approach, which is to “own some, but not all” of the apps used by consumers and end users of access services. No cable operator ever expects to own all or most of the relevant content being consumed by its video subscribers. But, where possible, it makes sense to own a few of those networks.

Some take matters even further. Liberty Global fully owns some assets--especially its access assets. But where it comes to app and content providers, owning parts of those operations, letting them run autonomously, and then participating in the equity upside, is a proven model.

The implications for former incumbent telcos, as their legacy revenue sources dwindle, is to cultivate a new approach based on maximizing total financial return, not return from internal branded services and assets.

In other words, owning--but not controlling assets--might provide a surer path to financial well being than controlling all the assets. To be sure, some mix of fully owned and minority investor models will likely be the path taken by most who can do so.

But the model of “owning access, but investing in apps” might be the most-logical way to create new asset value and revenue over the long term, compared to a strict “we own everything” model.

In that regard, the Internet of Things seems a promising area where partnerships, especially those where the access provider mostly has an ownership interest, but does not control IoT services and apps, seem logical.

For starters, there are simply too many market segments for a monolithic approach to work, too much domain knowledge required.

So even if “access” remains an obvious revenue generator, the volume of revenue still will lie in the apps. Better to own a small piece of that part of the business, than try and “own and control” very much of it.

Connectivity seems destined to a be a sliver of the total IoT market revenue.

source: Machina Research  

source: McKinsey Global Institute

By some estimates, there already are between six and 14 billion autonomous connected devices or “Things” that are connected via some form of communication mechanism, not including smartphones, tablets, computers and similar consumer devices. By way of comparison, there might be seven billion connected consumer devices in use today, including phones, PCs, tablets, TVs, game players and so forth.

As you might guess, the further out one goes, the more current estimates diverge. The number of connected devices by 2020 ranges between 18 billion and 50 billion devices. Some of us would argue the issue is not the number of devices in use, but the timeframe.

We will probably see less deployment up to 2020, then much more development after 2030.

source: IoT Analytics